Hi Federico,
It is possible, although it would take a small amount of munging from the
"Resources & Data" page of our website
(
http://rtw.ml.cmu.edu/rtw/resources).
To get the domains and ranges, you would need to look in the "ontology of
categories and relations" link. If you select only those rows where the
second column is "domain" or "range" then you will get a list of
(relation, domain/range, category) triples.
The bulk of what you want is found in the "extraction patterns learned by
CPL" link. If you consider only the first and third colums, then each
line gives you a single (predicate, extractionPatterns) pair. This will
include both categories and relations, so if you are interested only in
relations, then you can filter out everything that doesn't match the name
of one of the relations for which you know the domain and range from the
above step.
Let me know if that's unclear or isn't what you need.
bki...@cs.cmu.edu
> - extractionPatterns
> - "arg1 a college in arg2" "arg1 About car hire in arg2" "arg1
> about DD miles from downtown arg2" "arg1 Access Road to arg2" "arg1
> aerodrome near arg2".........
>
>
> Thank you for your attention
>
> King regards
> Federico
>
> --
> You received this message because you are subscribed to the Google Groups "NELL: Never-Ending Language Learner" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
cmunell+u...@googlegroups.com.
> For more options, visit
https://groups.google.com/d/optout.
>